Nan Li
I am a PhD candidate in Computational Linguistics and Natural Language Processing in the Natural Language Processing group at Utrecht University, advised by Prof. Massimo Poesio and Prof. Albert Gatt.
My research takes a perspectivist view of meaning: interpretation depends on the information and perspectives available to each interlocutor. I am broadly interested in dialogue, grounding, and how language and vision-language models represent meaning variation.
My work is part of DeMeVa and is funded by the Dutch Research Council (NWO) through the AiNed Fellowship Grant. I hold bachelor’s and master’s degrees in Computational Linguistics, both from Peking University, under the supervision of Prof. Weidong Zhan. I also interned at the Institute of Computational Linguistics at Peking University, working with Prof. Zhifang Sui.
news
| Jul 25, 2026 | I will present Seeing Is Not Sharing: Some Vision-Language Models Overestimate Common Ground in Asymmetric Dialogue at SIGDIAL 2026, Emory University, Atlanta, 3–5 August. My talk is in Oral Session 5 on 5 August. Come say hello if you are there. |
|---|---|
| Jul 20, 2026 | Next month I will be in Prague for HLV-DP (Human Label Variation in Discourse and Pragmatic phenomena) at ESSLLI 2026, 10–14 August, giving a long talk on 12 August: Personal Interpretations as Pragmatic Label Variation: Perspectivist Grounding Annotation and (V)LLM Evaluation. |
| Jun 01, 2026 | Seeing Is Not Sharing: Some Vision-Language Models Overestimate Common Ground in Asymmetric Dialogue has been accepted to SIGDIAL 2026 as an oral presentation. |
| May 14, 2026 | I presented Grounded Misunderstandings in Asymmetric Dialogue: A Perspectivist Annotation Scheme for MapTask at LREC 2026 in Palma de Mallorca. |
| Feb 13, 2026 | Grounded Misunderstandings in Asymmetric Dialogue: A Perspectivist Annotation Scheme for MapTask has been accepted to LREC 2026 in Palma de Mallorca. |
| Nov 08, 2025 | I presented our team submission DeMeVa at LeWiDi-2025 as a poster at the 4th Workshop on Perspectivist Approaches to NLP (NLPerspectives), co-located with EMNLP 2025 in Suzhou. |
selected publications
- Seeing Is Not Sharing: Some Vision-Language Models Overestimate Common Ground in Asymmetric DialogueOralIn Proceedings of the 27th Annual Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL), Aug 2026
- DeMeVa at LeWiDi-2025: Modeling Perspectives with In-Context Learning and Label Distribution LearningIn Proceedings of the 4th Workshop on Perspectivist Approaches to NLP (NLPerspectives), Nov 2025